Random-access memory

Random-access memory (RAM; ræm) is a form of computer memory that can be read and changed in any order, typically used to store working data and machine code. A random-access memory device allows data items to be read or written in almost the same amount of time irrespective of the physical location of data inside the memory, in contrast with other direct-access data storage media (such as hard disks, CD-RWs, DVD-RWs and the older magnetic tapes and drum memory), where the time required to read and write data items varies significantly depending on their physical locations on the recording medium, due to mechanical limitations such as media rotation speeds and arm movement. RAM contains multiplexing and demultiplexing circuitry, to connect the data lines to the addressed storage for reading or writing the entry. Usually more than one bit of storage is accessed by the same address, and RAM devices often have multiple data lines and are said to be "8-bit" or "16-bit", etc. devices.

Theories of topologically-induced phenomena in skyrmion-hosting magnetoelectric insulators

Oleksandr Kriuchkov

There once was a harsh competition between different computer memory technologies, and now we cheer triumph for the Random Access-Memory (RAM) devices, -- cheap, fast, tiny, stable. The competing Magnetic Bubble Memory had faded away as magnetic bubbles are undesirably large and pretty much not robust, manifesting both low data density and high operational costs. However, what we do possess now are nanosize objects of topological nature, magnetic skyrmions, which are protected from continuous field variations and take very little energy cost to be moved. Thus, skyrmions are considered as promising information carriers for future memory devices and ultradense data storage, while skyrmion phases in bulk materials are interesting from fundamental point of view in exploring topological states of matter. In this thesis, I develop and advance several effective theoretical approaches, diverse both in their methods and use, which were of appeal for several skyrmionic experiments in our lab (LQM/EPFL). We were and are primarily interested in the open issues in the applied field of skyrmionics, which may be taken under the umbrella of creation, stabilization and control of magnetic skyrmions under electric fields, mechanical strains, thermal gradients, etc. For the goals achieved and yet to be achieved, the magnetoelectric insulator Cu2OSeO3, which uniquely responses to all the above mentioned fields, is a highly advantageous candidate. Magnetoelectric here means that the spins of a magnetic material are coupled to external electric fields, while insulating properties are very advantageous to preserve both the state and the very existence of skyrmions by eliminating the Joule heating. The novel results in this thesis are calculations for both individual and arrayed skyrmions under electric fields, mechanical strains and uniform pressures, and thermal gradients. Furthermore, several fundamental questions were addressed by developing an extended formalism for calculation skyrmion-pocket phase diagrams, studying the topologically-governed crossover between skyrmions and magnetic bubbles, and discussing the possible role of merons (half-skyrmions) in skyrmion phase formation. The result with the most immediate appeal is probably the theoretical and experimental study of skyrmion lattices in electric fields, with a direct demonstration of writing and erasing of the full skyrmion phase under electric fields of few Volts per micrometer, as compatible with modern microelectronic devices.

EPFL2017

Rethinking Software Runtimes for Disaggregated Memory

Sanidhya Kashyap, Ivan Puddu

Disaggregated memory can address resource provisioning inefficiencies in current datacenters. Multiple software runtimes for disaggregated memory have been proposed in an attempt to make disaggregated memory practical. These systems rely on the virtual memory subsystem to transparently offer disaggregated memory to applications using a local memory abstraction. Unfortunately, using virtual memory for disaggregation has multiple limitations, including high overhead that comes from the use of page faults to identify what data to fetch and cache locally, and high dirty data amplification that comes from the use of page-granularity for tracking changes to the cached data (4KB or higher). In this paper, we propose a fundamentally new approach to designing software runtimes for disaggregated memory that addresses these limitations. Our main observation is that we can use cache coherence instead of virtual memory for tracking applications' memory accesses transparently, at cache-line granularity. This simple idea (1) eliminates page faults from the application critical path when accessing remote data, and (2) decouples the application memory access tracking from the virtual memory page size, enabling cache-line granularity dirty data tracking and eviction. Using this observation, we implemented a new software runtime for disaggregated memory that improves average memory access time by 1.7-5X and reduces dirty data amplification by 2-10X, compared to state-of-the-art systems.

ASSOC COMPUTING MACHINERY2021

Theories of topologically-induced phenomena in skyrmion-hosting magnetoelectric insulators

Oleksandr Kriuchkov

EPFL2017

Randomness Assists in Wireless Connectivity

Theories of topologically-induced phenomena in skyrmion-hosting magnetoelectric insulators

Rethinking Software Runtimes for Disaggregated Memory

Theories of topologically-induced phenomena in skyrmion-hosting magnetoelectric insulators

Randomness Assists in Wireless Connectivity

Rethinking Software Runtimes for Disaggregated Memory